GridPP Presentation to PPARC eScience Committee

1
GridPP Presentation to PPARC e-Science
Committee 31 May 2001
Steve Lloyd Tony Doyle
2
GridPP History
Collaboration formed by all UK PP Experimental
Groups in 1999 to submit 5.9M JIF bid for
Prototype Tier-1 centre at RAL (later withdrawn)
Added some Tier-2 support to become part of PPARC
LTSR - The LHC Computing Challenge ? Input to
SR2000
From Jan 2001 handling PPARCs commitment to EU
DataGrid
Formed GridPP in Dec 2000 included CERN, CLRC and
UK PP Theory Groups
3
Physics Drivers
Addresses one of PPARC's highest priority Science
programmes. LHC Experiments are the key to
understanding the origins of Mass (Higgs?),
Supersymmetry, CP Violation, Quark gluon plasma
and possible new phenomena e.g. extra dimensions
Maximise return from substantial UK investment in
LHC Detectors
4
The LHC Computing Challenge

• The problem is Huge
• Total data/year from one experiment 1 to 8 PB
  (Petabytes 1015 Bytes)
• Estimate total requirement to be
• 8M SI-95 of CPU Power (? 200,000 1 GHz
  PCs)
• 28 PB of 'Tape' storage
• 10 PB of disk storage

• The problem is Complex
• gt 108 electronic channels are read out each event
• The LHC will produce 8x108 pp interactions per
  second
• The Higgs to ?? rate for example is expected to
  be 2x10-4 per second
• A 2x10-4 needle in a 8x108 haystack

Distributed Solution to maximise use of
facilities and resources worldwide
5
LHC Computing Model
Lab m
Uni x
USA Brookhaven
Uni a
UK
Lab a
USA FermiLab
France
Tier 1
CERN
Uni n
Tier2
Physics Department
.
Italy
Desktop
Germany
NL
Lab b
Lab c
?
Uni y
?
Uni b
?
6
Proposal Executive Summary

• 40M 3-Year Programme
• LHC Computing Challenge
• Grid Technology
• Five Components
• Foundation
• Production
• Middleware
• Exploitation
• Value-added Exploitation
• Emphasis on Grid Services and Core Middleware
• UK computing strength within CERN

• Integrated with EU DataGrid, PPDG and GriPhyN
• Facilities at CERN, RAL and up to four UK Tier-2
  sites
• Centres Dissemination
• LHC developments integrated into current
  programme (BaBar, CDF, D0, ...)
• Robust Management Structure
• Deliverables in March 2002, 2003, 2004

7
GridPP Component Model
Value-added
Component 5 Value-Added Full exploitation of
Grid potential for Particle Physics
25.9M
Exploitation
Component 4 Exploitation The applications
necessary to deliver Grid based Particle Physics
21.0M
Middleware
Component 3 Middleware Connecting the Foundation
and the Production environments to create a
functional Grid
17.0M
Production
Component 2 Production Built on Foundation to
provide an environment for experiments to use
12.7M
Component 1 Foundation The key infrastructure at
CERN and within the UK
Foundation
8.5M
8
Major Deliverables

• Prototype I - March 2002
• Performance and scalability testing of components
• Testing of the job scheduling and data
  replication software from the first DataGrid
  release.

• Prototype II - March 2003
• Prototyping of the integrated local computing
  fabric, with emphasis on scaling, reliability and
  resilience to errors.
• Performance testing of LHC applications.
  Distributed HEP and other science application
  models using the second DataGrid release.

• Prototype III - March 2004
• Full scale testing of the LHC computing model
  with fabric management and Grid management
  software for Tier-0 and Tier-1 centres, with some
  Tier-2 components.

9
Financial Summary
Components 1-4 PPARC External Funds UK
Staff 10.7M 5.9M (EPSRC?) UK Capital 3.2M
4.5M (SRIF?) CERN Staff 5.7M CERN Capital
1.4M Total 21.0M 10.3M
Computing Science
LHC Tier-0
Up to 4 Tier-2
LHC Tier-1/BaBar Tier A
10
GridPP Organisation
Hardware development organised around a number
Regional Centres

• Likely Tier-2 Regional Centres
• Focus for Dissemination and Collaboration with
  other disciplines and Industry
• Clear mapping onto Core Regional e-Science Centres

Software development organised around a number of
Workgroups
11
GridPP Workgroups
Technical work broken down into several
workgroups - broad overlap with EU DataGrid
A - Workload Management Provision of software
that schedule application processing requests
amongst resources
F - Networking Network fabric provision through
to integration of network services into middleware
G - Prototype Grid Implementation of a UK Grid
prototype tying together new and existing
facilities
B - Information Services and Data
Management Provision of software tools to provide
flexible transparent and reliable access to the
data
H - Software Support Provide services to enable
the development, testing and deployment of
middleware and applications at institutes
C - Monitoring Services All aspects of monitoring
Grid services
I - Experimental Objectives Responsible for
ensuring development of GridPP is driven by needs
of UK PP experiments
D - Fabric Management and Mass Storage Integration
of heterogeneous resources into common Grid
framework
E - Security Security mechanisms from
Certification Authorities to low level components
J - Dissemination Ensure good dissemination of
developments arising from GridPP into other
communities and vice versa
12
GridPP and CERN
For UK to exploit LHC to the full Requires
substantial investment at CERN to support LHC
computing.

• UK involvement through GridPP will boost CERN
  investment in key areas
• Fabric management software
• Grid security
• Grid data management
• Networking
• Adaptation of physics applications
• Computer Centre fabric (Tier-0)

13
GridPP and CERN
This investment will

• Allow operation of a production quality prototype
  of the distributed model prior to acquisition of
  final LHC configuration
• Train staff for management and operation of
  distributed computing centres
• Provide excellent training ground for young
  people
• Enable the technology to be re-used by other
  sciences and industry

14
Staff at CERN
Proposal is that staff are hired by UK
Universities or Laboratories and sent on
long-term mission to CERN. Employed as CERN
associates in IT Division

• Integrated part of the CERN LHC activity
• Staff assigned to CERN teams with operational and
  development responsibilities
• Each team responsible for LHC development and
  prototyping as well as for operating current
  services
• Developers need hands-on operational experience
• Ensures that CERN experience is fully utilised
• Formal LHC Computing Project structure being
  defined, to ensure overseeing role by funding
  bodies to be agreed with CERN Council

15
Hardware at CERN
16
GridPP Management Structure
17
Management Status
The Collaboration Board (CB) The governing body
of the project - consists of Group Leaders of all
Institutes - established and Collaboration Board
Chair elected
The Peer Review Selection Committee
(PRSC) Pending approval of Project
The Project Management Board (PMB) The Executive
board chaired by Project Leader - Project Leader
being appointed. Shadow Board in operation
The Dissemination Board (DB) Pending approval of
Project
The Technical Board (TB) The main working forum
chaired by the Technical Board Chair - interim
task force in place
The Experiments Board (EB) The forum for
experimental input into the project - nominations
from experiments underway
18
Information Flow
19
Meetings Schedule
Yearly Reporting to PPARC?
Quarterly Reporting to EU
20
GridPP Collaboration Meeting
1st GridPP Collaboration Meeting - Coseners House
- May 24/25 2001
21
UK Strengths

• Wish to build on UK strengths -
• Information Services
• Networking - world leaders in monitoring
• Security
• Mass storage

• UK Major Grid Leadership roles -
• Lead DataGrid Architecture Task Force (Steve
  Fisher)
• Lead DataGrid WP3 Information Services (Robin
  Middleton)
• Lead DataGrid WP5 Mass Storage (John Gordon)
• ATLAS Software Coordinator (Norman McCubbin)
• LHCb Grid Coordinator (Frank Harris)

• Strong UK Collaboration with Globus
• Globus people gave 2 day tutorial at RAL to PP
  community
• Carl Kesselman attended UK Grid Technical
  meeting
• 3 UK people visited Globus at Argonne

Natural UK Collaboration with US PPDG and GriPhyN
22
Funding Requirements
Full exploitation requires 25.9M from PPARC plus
11.6M external funds
Minimum programme requires 21.0M from PPARC plus
10.3M external funds
Profiling is driven by Hardware Immediate need
(Buy now) v Moore's law (Buy later) ? UK Flat,
CERN rising Manpower Immediate need
requirement for 3 year positions (Hire now) v
Availability (spread out) ? UK CERN want to
front load Does not match PPARC funding profile
Our proposal profiled this as 2001/2 2002/3 2003
/4 3.91M 8.43M 8.64M c.f. PPARC
profile 3.0M 8.0M 15M
23
Proposal Profile
24
Reprofiling
One attempt to match PPARC profile
Too many staff hired in 3rd year (for 1 year!)
25
First Year Deliverables
Each Workgroup has detailed deliverables. These
will be refined each year and build on the
successes of the previous year.

• The Global Objectives for the first year are
• Deliver EU DataGrid Middleware (First prototype
  M9)
• Running experiments to integrate their data
  management systems into existing facilities (e.g.
  mass storage)
• Assessment of technological and sociological Grid
  analysis needs
• Experiments refine data models for analyses
• Develop tools to allow bulk data transfer
• Assess and implement metadata definitions
• Develop relationships across multi-Tier
  structures and countries
• Integrate Monte Carlo production tools
• Provide experimental software installation kits
• LHC experiments start Data Challenges
• Feedback assessment of middleware tools

26
External Resources
External Funds (additional to PPARC Grants and
central facilities) have provided computing
equipment for several experiments and institutes

• BaBar (Birmingham, Bristol, Brunel, Edinburgh,
  0.8M (JREI) 1.0M (JIF)
• Imperial, Liverpool, Manchester, QMUL, RAL,
  RHUL)
• MAP (Liverpool ) 0.3M (JREI)
• ScotGrid (Edinburgh, Glasgow) 0.8M (JREI)
• D0 (Lancaster) 0.4M (JREI) 0.1M (Univ)
• Dark Matter (Sheffield) 0.03M (JIF)
• CDF/Minos (Glasgow, Liverpool, Oxford, UCL) 1.7M
  (JIF)
• CMS (Imperial) 0.15M (JREI)
• ALICE (Birmingham) 0.15M (JREI)
• Total 5.4M

All these Resources will contribute directly to
GridPP
Many Particle Physics Groups are involved in
large SRIF bids in collaboration with other
disciplines mostly to form e-Science centres. The
amount of resource available to GridPP from this
SRIF round could be several M
27
First Year Priorities

• Funding of PPARC's EU DataGrid staff commitment
• Staff to implement initial Grid Testbed in UK
• Hardware to satisfy BaBar requirements and EU
  DataGrid testbed commitment
• Staff for CERN LHC computing
• Contribution to CERN Tier-0

(Staff costs assume only 6 months of salary in
first year)

• 15 EU DataGrid 3 year posts (already committed)
  0.4M
• 15 GridPP 3 year posts 0.4M
• Hardware for BaBar Tier-A/LHC Tier-1 0.9M
• 15 CERN 3 year posts 0.5M
• Hardware for CERN Tier-0 0.3M
• Total 2.5M

Minimum viable programme to meet commitments
28
Summary

• Have been working towards this project for 2
  years building up hardware
• Balanced exploitation programme costing 21M
• Will put PPARC and the UK at the forefront of
  Grid development in Europe
• Funds installation and operation of experimental
  testbeds, key infrastructure, generic middleware
  and making application code grid aware
• Does NOT fund physics analysis or experiment
  specific algorithms
• Does not match well with PPARC's funding profile
• A mechanism for moving some money forward from
  the third year needs to be found
• Requires substantial investment NOW
• 2.5M required this year to kick start programme